This invention relates generally to the field of electrostatic speakers and more specifically to power supplies for biasing electrostatic speakers.
The long history of electrostatic speakers has produced a wide variety of speaker configurations. To provide a linear output, an electrostatic speaker requires a high (500 to 5000 volts) substantially DC (direct current) voltage to be applied either to the stators or the diaphragm. This applied voltage creates a DC constant for the AC (alternating current) signal voltages to work against. Since only the leakage currents need to be supplied, the wattage rating of the fixed bias supply can be quite low (less than a watt) and the package size can be small (a few cubic inches).
Historically, this DC voltage has been provided by running a step-up transformer from an AC power line, rectifying its output, and connecting the rectified output to a capacitor. U.S. Pat. No. 2,896,025 granted to Janszen embodies this approach. This configuration is easy to implement but can be somewhat costly. It can also be inconvenient to have to run separate AC main wires and also signal wires from the power amplifier. Additionally, if the AC power is intended to be supplied directly from a wall source, there may be no AC power sockets located nearby the electrostatic loudspeakers. Another drawback of using a separate AC power supply is that the separate power supply results in additional cost and wiring which makes electrostatic speakers a less desirable choice in most consumer applications. Thus, the electrostatic speakers are less desirable even though they offer superior performance and greater sound fidelity when they couple into the air.
In particular applications where the systems run off of DC, such as a laptop computer or a portable music system, a high voltage source of AC may not be available. In these applications, a DC to DC converter is required to produce the required high voltages. This DC to DC convertor system is illustrated in U.S. Pat. No. 3,992,585 granted to Turner, et al.
Another method to provide a DC bias, which avoids many of the issues in the prior art listed above, is to tap off of the secondary winding of the audio signal transformer. The tapped voltage is then rectified and the energy is stored in a capacitor. Because the bias currents are near zero, this approach has virtually no impact on the signal currents. Disclosures of this technique can be found in U.S. Pat. No. 3,895,193 granted to Bobb, U.S. Pat. No. 4,160,882 granted to Driver and U.S. Pat. No. 5,392,358 granted to Driver.
For most consumer applications, what would be most useful, is a xe2x80x9cdrop inxe2x80x9d replacement for existing electromagnetic speakers. In other words, an electrostatic speaker which can effectively replace existing electromagnetic speaker systems is desirable. This would eliminate the need for an AC outlet or a DC to DC convertor and maintain a simple connection with two wires for each speaker. Self-biasing can provide this, but the prior art systems all suffer from a common group of significant drawbacks.
First, because the AC audio signal is not predictable or repeatable, the voltage available at the output of the audio signal step-up transformer can vary from a zero voltage to a voltage that can damage the electrostatic unit due to over voltage.
A second problem with the prior art type of bias system is that when the audio equipment is first powered up, the self-bias voltage (and hence the resulting electric field) is at, or close to zero. As a result, there is a start up time during which the audio level gradually increases to the maximum. During the charging period, the program signal will not be heard at its proper volume. For certain types of music and some audio material, many seconds elapse before the self-bias voltage comes into its normal range. One approach is to have a fast signal rise time when the system is turned on. To increase the signal rise time, the transformer step-up ratio can be increased but this can then make the first problem of over-voltage even worse.
A third problem is that prior art self-bias circuits provide a variable bias voltage. The side effect of the variable bias voltage can best be described as producing a noticeable xe2x80x9cpumping actionxe2x80x9d in the reproduced acoustic output level.
A fourth problem with this type of bias system is that in a multi-channel system, each channel can end up with different bias levels at any given time. Therefore, each channel would have a different efficiency and would be mismatched depending on how well the multi-channel program material was matched from channel to channel at any given moment.
It is an object of the present invention to provide a bias system which uses a simple two wire speaker connection and the reduces cost of self-biasing for DC field generation in the electrostatic speaker.
It is also an object to provide a bias system which uses an audio signal derived bias system for an electrostatic loudspeaker that steps up the voltage to the required level even with program material at a low level, and maintains a substantially constant supply voltage.
It is another object of the current invention to provide a self bias system which uses a voltage limiting/regulating means in an output signal fed bias supply so the voltage is stabilized to be substantially constant and limited from over voltage.
It is a further object of the invention to achieve a more effective startup than prior art systems using the greater step-up ratio of the transformer secondaries.
It is an additional object of the invention to provide a more effective startup using greater multiplication stages in the voltage multiplier circuit and a separate charging signal delivered from the associated active electronics which charges on startup, periodically, or on a steady basis.
The presently preferred embodiment of the present invention is an audio signal derived bias supply for use with an electrostatic loudspeaker. The bias supply includes at least one transformer adapted to receive an audio signal. The transformer has at least one primary winding, and primary connection taps. The transformer also has at least one secondary winding magnetically coupled to the primary winding, which has at least two secondary connection taps. A bias circuit is connected to the at least one secondary winding. The bias circuit has a rectification means and a voltage limiting means, coupled to the rectification means.
These and other objects, features, advantages and alternative aspects of the present invention will become apparent to those skilled in the art from a consideration of the following detailed description taken in combination with the accompanying drawings.